Airport Delay Hotspots: Which Airports Have the Worst Delays

Some airports have far more delays than others, and it’s rarely bad luck. Their layout, their traffic and their weather produce delays season after season. A hub handling five hundred movements a day works under very different limits from a small point-to-point airport, and the numbers show it.

Below: why delays concentrate at certain airports, how to read the public data that tracks them, and how event contracts let traders and travelers take a position on delay risk before a disruption happens. It includes 2026 figures for US and European airports from published US Bureau of Transportation Statistics (BTS) and Eurocontrol data.

Why Airports Become Delay Hotspots

Hub Congestion and Slot Constraints

Major hubs handle hundreds of aircraft movements an hour. Level 3 slot-coordinated airports are a designation under EU Council Regulation 95/93, used where demand regularly exceeds declared capacity. At these airports a neutral coordinator allocates take-off and landing slots in advance. High-traffic European hubs such as London Heathrow, Paris Charles de Gaulle and Frankfurt Airport work this way, and every slot is filled, so there’s no spare capacity to absorb a delay.

When an inbound aircraft arrives late, the turnaround gets shorter and the outbound departure takes on the delay. At a hub with tight rotations, one late inbound can delay dozens of onward connections. Ground handling is stretched too. During peak departure banks, ramp crews, fuelers and catering are scheduled for maximum throughput. Any slip in the schedule causes clashes over staff and equipment, and turnaround times grow across the whole pier.

Weather as a Structural Factor

Weather delays aren’t spread evenly across airports. Some airports sit where the same weather comes back every year. Coastal airports get advection fog in spring and early summer. Inland airports get thunderstorms in the peak summer months. Northern airports need de-icing, and demand for it rises steeply once temperatures fall below freezing.

That makes the seasonal pattern predictable. Airports in thunderstorm-prone regions tend to show higher delay rates in Q2–Q3, however well their ground handling and airlines run. A “worst airports” ranking published in January looks very different from the same data in August. Delay statistics are seasonal, and reading them without the month attached is a common mistake.

ATC Restrictions and Flow Management

Air traffic flow management (ATFM) is how the system copes when demand exceeds declared airspace capacity. In Europe, Eurocontrol’s Network Manager Operations Centre issues ATFM regulations that keep aircraft on the ground at the departure airport instead of letting them circle in the air. That avoids airborne holding. It also means the delay shows up at departure, and it’s attributed to the destination airport’s capacity limits, even though the aircraft is waiting at the origin.

In the United States, the Federal Aviation Administration (FAA) uses ground delay programs (GDPs) and ground stops for the same purpose. Airports that are often under ATFM restrictions build up delays that persist across airlines and seasons. The limit is the infrastructure, and no single airline can route around it.

The Connectivity Ripple Effect

Hubs have high rotation density: many movements for each aircraft based or operating there. The same aircraft flies several legs a day. A delay on its first leg carries into every later leg, even legs that never touch the same hub. So a hub’s delay figures include its own constraints plus delays that started anywhere in its inbound network.

That’s why the first departure of the morning, after the aircraft has sat overnight, is more punctual than afternoon and evening flights on the same route. By midday, delays from across the network show up in the on-time figures.

For travelers connecting through a major hub, this ripple raises missed connection risk, and the risk grows with rotation density. Tight minimum connection times add to it. Minimum Connection Time: The Hidden Missed-Connection Risk breaks this down by route.

How to Read Airport Delay Data

Key Metrics Explained

On-time performance (OTP) is the standard metric: the share of flights that depart or arrive within 15 minutes of schedule. BTS uses this 15-minute threshold for US domestic reporting, and Eurocontrol’s Central Office for Delay Analysis (CODA) uses it widely in its reports on European flights.

Two directional measures matter in practice:

• ADEP (aerodrome of departure) delay: delay attributed to the departure airport, covering airline, ground handling and ATC ground-hold causes.
• ADES (aerodrome of destination) delay: delay attributed to en-route and destination factors, including airspace restrictions and arrival-airport capacity limits.

IATA delay codes (numeric codes 11–99) record the cause of each delay under IATA’s standard delay coding system (AHM 730), which Eurocontrol CODA uses in its delay-cause reporting. Codes 11–48 cover airline and handling delays (passenger and baggage handling, cargo, aircraft and ramp handling, and technical causes). Codes 61–69 cover flight operations and crewing. Codes 71–77 cover weather (departure-airport weather, destination weather, en-route or alternate conditions, de-icing, and ground handling slowed by weather). Codes 81–83 cover ATC/ATFM restrictions (en-route capacity, ATC staff and equipment, and arrival-airport restrictions). The cause mix matters as much as the headline OTP figure. A 30% delay rate driven by ATC codes (81–83) points to infrastructure limits. The same rate driven by airline and handling codes (11–48) points to how the airline and ground staff perform.

Public Data Sources

These are the main public sources. The 2026 figures they give are in the data sections below.

• BTS: monthly on-time performance data for US domestic flights, free on the TranStats portal, with raw per-flight files and IATA-coded causes.
• Eurocontrol performance data: the Eurocontrol performance data portal publishes airport-level ATFM delay datasets each month, plus CODA digests that break down delay causes.
• FAA: the National Airspace System status page shows active flow restrictions in real time. The FAA’s air traffic pages explain how those restrictions work.
• Operator publications: many large airports and airlines publish monthly punctuality and traffic figures in operating statistics or investor updates. These cover flights the national datasets above don’t.
• The airport’s own board: every major airport publishes live arrivals and departures with a scheduled and an actual time for each flight. It’s the most detailed public record there is and it’s free. On GADUIN, it’s the record a flight market settles on.

When you quote an OTP figure or ranking, check the publication and the period it covers. Last year’s figure doesn’t tell you how an airport is doing now.

What “Worst Airport” Rankings Actually Measure

Different methods give different results, and consumer rankings often mix them up. The main variables:

Absolute count vs. share: an airport with 600,000 movements a year will have more delayed flights than one with 60,000, even if its OTP percentage is better. Rankings based on raw delay counts put large hubs at the top regardless of how well they run.

Domestic vs. international scope: BTS data covers US domestic flights only. Eurocontrol data covers its network area and, as the next section shows, measures something else entirely. A ranking built on one can’t be compared directly with a ranking built on the other.

Cause attribution: rankings that leave out weather or ATC delays, usually described as “factors outside the carrier’s control”, put airports in a different order from all-cause rankings. There’s no universal standard.

Airport-level vs. route-level: an airport’s overall OTP can look fine while one busy route through it is late again and again. Travelers on that route get the route’s delay rate, whatever the airport average is.

Broad “worst airports” lists are good for a general picture. For a specific trip or hedge, route-level data is more useful: from the national datasets, or from the airport’s own board for the flight you’re actually booked on. The two sections below apply this to the latest BTS and Eurocontrol figures.

Worst US Airports for Delays: A 2026 Data Snapshot

The latest published month of BTS data is May 2026, from the BTS TranStats on-time performance portal and its raw monthly reporting file. It uses the standard BTS method: the 15-minute arrival threshold above, with cancelled and diverted flights counted in the denominator. Average arrival delay is per completed arrival, with early and on-time flights counted as zero. The ranking covers airports with at least 6,000 scheduled arrivals in the month, worst first.

Airport	On-time arrivals	Avg arrival delay (min)	Cancelled
San Francisco International (SFO)	53.3%	30.0	0.8%
Dallas Love Field (DAL)	68.6%	18.7	1.2%
Nashville International (BNA)	71.9%	16.7	0.6%
Austin–Bergstrom International (AUS)	72.5%	17.6	0.7%
San Diego International (SAN)	73.0%	15.0	0.6%
Chicago Midway International (MDW)	73.7%	14.4	0.5%
Ronald Reagan Washington National (DCA)	74.1%	18.9	1.4%
Dallas Fort Worth International (DFW)	74.2%	23.8	3.4%
Fort Lauderdale–Hollywood International (FLL)	74.5%	16.8	0.5%
Tampa International (TPA)	74.5%	17.2	0.6%

BTS data for May 2026. Across the system, the month had 611,735 scheduled domestic flights, 77.6% on-time arrivals and a 0.9% cancellation rate.

San Francisco is the outlier, more than 24 percentage points below the system average. 64% of its attributed delay minutes fall in the National Aviation System (NAS) category, which covers airspace, air traffic control and capacity causes, as opposed to airline operations. That fits SFO’s closely spaced parallel runways, which lose arrival capacity when visibility drops. Dallas Fort Worth looks different. It had a 3.4% cancellation rate, the highest among major airports that month, and its delays were led by late-arriving aircraft (37%) and carrier causes (31%). Keep in mind that May’s ranking reflects May’s weather and traffic. The seasonality section below covers this.

Europe’s Delay Hotspots: What Eurocontrol Data Shows

European figures can’t be compared directly with the BTS table above, because Eurocontrol measures something else. Its Airport Arrival ATFM Delay dataset records the minutes of air traffic flow management delay attributed to each arrival airport, averaged per arrival. These are ground holds put on inbound flights when a regulation caps the airport’s arrival rate. A flight can push back 40 minutes late and still show zero ATFM delay. BTS counts how far a flight ran from its schedule at the gate. Eurocontrol counts flow delay imposed by regulations. Keep the two apart.

The latest published month in the dataset is June 2026. Among airports in the Eurocontrol network area with at least 5,000 arrivals that month, these had the highest average arrival ATFM delays:

Airport	Avg arrival ATFM delay (min/arrival)	Dominant recorded cause
Tel Aviv Ben Gurion (LLBG)	9.4	ATC capacity (~100%)
Nice-Côte d’Azur (LFMN)	6.6	ATC capacity (98%)
Athens (LGAV)	4.4	ATC capacity (82%)
Zürich (LSZH)	3.4	Other (75%), weather (15%)
London Gatwick (EGKK)	2.6	Weather (78%)
Porto (LPPR)	1.8	Weather (83%)
London Heathrow (EGLL)	1.8	Weather (52%), capacity (43%)
Amsterdam Schiphol (EHAM)	1.6	Aerodrome capacity (78%)
Lisbon (LPPT)	1.3	Capacity (69%), weather (28%)
Barcelona (LEBL)	1.1	Mixed: weather (30%), capacity (29%)

Eurocontrol data for June 2026. The dataset covers the whole Eurocontrol network area, which reaches beyond geographic Europe. That’s why Tel Aviv is on the list.

Across the network, June 2026 averaged 0.71 minutes of airport-attributed arrival ATFM delay per arrival. Capacity caused 56% of the delay minutes and weather 28%. The cause column is the one that matters for trading. Capacity-limited airports such as Nice or Athens produce delays whenever summer demand runs into a fixed arrival rate. Weather-driven airports such as Gatwick or Porto rise and fall with the forecast.

Seasonality: When a Hotspot Is Actually Hot

Annual rankings squeeze twelve very different months into one number and hide most of the useful information. The Eurocontrol dataset above lets you measure the seasonal spread. Across calendar 2025, network-wide airport-attributed arrival ATFM delay ran from 0.39 minutes per arrival in February to 1.09 in July, nearly three times as much within one year. Eurocontrol’s 2025 annual review of European aviation shows the same summer peak across the network.

The causes change with the calendar as much as the totals do. In January 2026, weather caused 77% of the network’s airport-attributed ATFM delay minutes. Fog, snow and de-icing dominate when traffic is light. By June 2026 the weather share had fallen to 28%, and capacity took over as schedules filled up toward the summer peak. Winter and summer hotspots are often different airports. Northern airports limited by de-icing top the January lists, and slot-saturated Mediterranean holiday airports show up in July.

So when you take a position, use the base rate for the month you’re trading. An airport in the middle of a full-year ranking can be in the top 10% for delays in its bad season. The seasonal flight delay trading strategy guide shows how calendar patterns turn into trading windows. The fog versus thunderstorm delay profile comparison explains why the two main weather causes produce differently shaped delays.

Why “Just Avoid Bad Airports” Isn’t a Risk Strategy

Knowing that a hub has high delay rates is useful. On its own, it doesn’t manage any risk.

Corporate travel programs, tour operators and business travelers usually can’t avoid the major hubs. The connecting hub depends on the city pair, the airline’s network and which fares are available. If that hub is a known delay hotspot, you’re going through it anyway.

The usual remedies come after the fact. The EU261 compensation framework lets you claim after a disruption on qualifying flights departing from Europe. It doesn’t prevent the delay, and it doesn’t cover the cost of a missed meeting or the knock-on costs of rebooking in peak season.

Knowing a hub gets delays in summer thunderstorm season tells you the chance of a delay is higher. It doesn’t give you anything to offset that risk ahead of time. For that you need a different instrument.

Turning Hotspot Data into a Contract Thesis

Rankings become useful once you turn them into a base rate. If BTS reports that an airport had 53.3% of arrivals on time, a simple base rate for any one arrival there is roughly a 47% chance of arriving 15 minutes late or worse, or being cancelled. That’s a starting point. Three refinements turn it into a trade idea.

First, match the threshold. GADUIN flight markets count a flight as Delayed when it arrives more than 15 minutes late, the same line as the BTS 15-minute arrival metric, so the BTS on-time share maps across directly. Averages can still mislead. An airport with a 30-minute average arrival delay that comes from a few very bad days has a very different share of flights past 15 minutes from one with the same average spread evenly.

Second, use the calendar. An annual base rate gets both the busy season and the quiet one wrong. The same month in previous years is a better anchor, adjusted for structural changes such as a runway closure, a bigger schedule or a new flow regulation.

Third, compare the base rate with the market price. A well-known hotspot is well known. A GADUIN price is the chance the market gives that outcome right now, and it got there because people read the same published statistics you do. You only have a reason to trade where your estimate differs from that price. That might be because you price the season more precisely, you account for a structural change the averages miss, or you spot a short-term signal such as an active flow program. Where your estimate and the price agree, there’s no edge, however bad the airport. A base rate is a probability, and it promises nothing about a single flight.

How to Hedge Airport Delay Risk with Event Contracts

What GADUIN Event Contracts Are

GADUIN runs event contracts on transport outcomes: whether a given flight arrives On time, is Delayed past the market’s threshold, or is Cancelled. Each of the three outcomes has a live price between 1¢ and 99¢, which is the chance the market gives it. Shares on the outcome that happens pay $1 in USDT, and the others pay $0. It isn’t insurance. There’s no claims process, no policy document and no underwriter. The market settles automatically on the destination airport’s published arrival time, compared with its published schedule and a threshold that’s fixed when the market opens and never changes.

For the full mechanics and settlement rules, see How Flight Delay Event Contracts Work on GADUIN.

From Data to Decision

Published national and network delay data is the research behind a position in a flight delay market. If a hub’s published figures show high delay rates in the summer peak, that history shapes your own estimate of the chance of On time, Delayed or Cancelled on a specific route. Those monthly releases are research only. A GADUIN flight market settles flight by flight, on the destination airport’s published arrival time against its published schedule.

A trader or corporate travel manager who has studied route-level OTP data has a basis for trading a specific flight. Buy Delayed if the delay risk looks higher than the market price suggests, or On time if the market seems to overprice a delay. The Hedge Your Flight Delay guide covers both directions in detail.

Who Benefits

Retail traders and travelers can trade a specific upcoming flight and hold a position on the delay outcome, with no claim to file afterwards. If the flight arrives later than the market’s threshold, the Delayed shares pay out in USDT. There’s no paperwork, no waiting period and no one reviewing a claim.

Corporate travel managers and institutional hedgers who book the same routes through delay-prone hubs again and again can use event contracts to offset the cost of delays: missed connections, rebooking and hotel nights, across many trips. Instead of a claims process, the market settles automatically on the airport’s published arrival time, under terms fixed before the market opened.

Frequently Asked Questions

Which US airport has the worst delays in 2026?

In the latest published BTS month, May 2026, San Francisco International had the lowest on-time arrival share among major US airports at 53.3%, against a system average of 77.6%. Most of its delay came from National Aviation System (airspace and capacity) causes.

Which European airports have the highest ATFM delays?

In Eurocontrol’s June 2026 data, Tel Aviv Ben Gurion (9.4 minutes of arrival ATFM delay per arrival), Nice (6.6) and Athens (4.4) led the network among airports with at least 5,000 arrivals in the month. At all three, ATC capacity was the main cause.

Can BTS and Eurocontrol delay statistics be compared directly?

No. BTS measures whether a flight reached the gate within 15 minutes of schedule. Eurocontrol measures minutes of flow-management ground hold attributed to the arrival airport. The same airport can look fine on one and strained on the other.

How current are airport delay rankings?

BTS publishes monthly, about two months behind. As of August 2026, May is the latest month available. Eurocontrol’s airport ATFM delay dataset also updates monthly, and June 2026 is the latest published period.

Does a hotspot ranking mean flights there will be delayed?

No. A ranking describes base rates across thousands of flights. It can’t tell you what one departure will do. A specific flight’s outcome depends on its route, time of day, season, and the weather and flow restrictions on the day. That’s why event contracts settle on the recorded outcome of one flight.

Airport delay statistics change with network changes, infrastructure investment, seasonal weather and regulation. The US figures above are BTS data for May 2026, and the European figures are Eurocontrol data for June 2026, the latest published periods at the time of writing. Check current data before you use them. Trading event contracts on GADUIN carries market risk, and outcomes aren’t guaranteed. This content is for information only and isn’t financial or investment advice. US persons can’t take part.